FDA’s Process Validation guidance calls for continuous process verification.

Here’s how to do it, and how automation can help

In January of 2011, the FDA issued Guidance for The guidelines identify three key program elements for CPV:
Industry, Process Validation: General Principles and 1. A system or systems for detecting unplanned depar-
Practices. This document will affect how pharmaceutical tures from normal operation of the process designed to
manufacturers operate and presents them with a series of correct, anticipate and prevent problems.
both challenges and opportunities. 2. A n ongoing program to collect and analyze prod-
The Guidance document, in its own words, “aligns uct and process data that relate to product quality,
process validation activities with a product lifecycle including evaluation of intra-batch and inter-batch
concept and with existing FDA guidance, including variation. This data “should include relevant process
the FDA/International Conference on Harmonization trends and quality of incoming materials or compo-
(ICH) guidances for industry, Q8(R2) Pharmaceutical nents, in-process material and finished products. The
Development, Q9 Quality Risk Management, and Q10 data should be statistically trended and reviewed by
Pharmaceutical Quality System. Although this guidance trained personnel. The information collected should
does not repeat the concepts and principles explained verify that the quality attributes are being appropri-
in those guidances, FDA encourages the use of modern ately controlled throughout the process.”
pharmaceutical development concepts, quality risk 3. Maintenance of the facility, utility and equipment
management, and quality systems at all stages of the qualification status: “Once established,” say the Guide-
manufacturing process lifecycle.” lines, “qualification status must be maintained through
The document defines process validation as “the routine monitoring, maintenance, and calibration
collection and evaluation of data, from the process design procedures and schedules.”
stage through commercial production, which establishes
scientific evidence that a process is capable of consistently Alignment with ICH Q10
delivering quality product.” Validation activities are ICH Q10, Pharmaceutical Quality System (June 2008)
broken into three stages: Process Design, Process is a tripartite guideline that describes “a model for an
Qualification and Continued Process Verification. effective quality management system for the pharmaceu-
This article will discuss the third stage, Continued tical industry, referred to as the Pharmaceutical Quality
Process Verification (CPV), the challenges and System.” One of its major objectives is to establish and
opportunities that this concept creates for pharmaceutical maintain a state of control: “To develop and use effective
companies, and what this means to pharmaceutical monitoring and control systems for process performance
manufacturers on a practical, day-to-day basis. and product quality, thereby providing assurance of
P r o c e s s Va l i d at i o n, A u t o m at i o n a n d C o n t r o l

continued suitability and capability of processes.” To a incorporated into the automation strategy to respond to
large extent ICH Q10 embodies existing regional GMP and correct process deviations in real time.
requirements. Its coverage extends from development to • Process monitoring — A DCS collects a consider-
manufacturing to product discontinuation. able amount of process data on an ongoing basis that
Several key elements of ICH Q10 align with Continued includes critical process parameters from moment-to-
Process verification as it relates to the “Manufacturing” moment loop control operations via reporting of trends.
stage. Two of the most important are: Additional process data may be generated outside of
Manufacturing through Quality Control testing.
• K nowledge Management, which ICH Q10 considers one Effective monitoring of the process requires both sets
of its two key enablers (the other being Risk Manage- of data. Integration of an EBR with Laboratory Informa-
ment) and is defined as a “systemic approach to acquir- tion Management Systems (LIMS) would provide
ing, analyzing, storing and disseminating information a central repository for data when test results are
related to products, processes and components.” The provided back to the batch record. Potential delays in
data historians found in modern distributed control obtaining laboratory data should be considered as they
systems (DCSs), often in combination with an electronic affect the ability to respond to results in real time. Iden-
batch record system, perform this function. tification of PAT with online testing in key areas would
• Continual improvement of process performance and contribute to the success of continued process verifica-
product quality, described in Section 3 of ICH Q10. tion where a response can occur in real time.
Regarding the commercial manufacturing stage of the
product lifecycle, A final data set
ICH Q10 states “the Make sure that the process control is required for raw
pharmaceutical strategy is appropriate, and that material information
quality system from sources such as
process, reporting and quality
should assure that certificates of analysis
the desired product parameters are identified/assigned. from suppliers (COAs)
quality is routinely and/or site release
met, suitable process performance is achieved, the set testing. If this data is collected in an ERP and/or LIMS
of controls are appropriate, improvement opportunities system, it can also be integrated into a central repository
are identified and evaluated, and the body of knowledge to provide ease of use for process analysis.
is continually expanded.”
• Maintenance of facility, utility and equipment — The data
In order to meet many, if not most of the goals of all handled by a DCS can be used to monitor the health of the
these guidelines, data collection and availability should equipment and process. Alarm records and key utility and
be consistent and easily accessible to users. These goals equipment performance parameters, as well as data from
can be achieved using current automation and business intelligent field devices, can be collected and analyzed. All
technologies that include Distributed Control Systems (DCS) of this data can be correlated with asset management and
and Manufacturing Execution Systems (MES). For example: MES systems to aid in predictive/proactive maintenance.

• Handling unplanned departures — A modern DCS As explained above, facilities that use modern distributed
has an alarm strategy and can be configured with control systems already have access to the data required
preplanned actions to respond to unwanted changes in for Continued Process Verification. But in many cases,
continuous data (e.g. temperature of a bioreactor). The these systems collect an enormous amount of data that
system can provide exception reporting through its is useful in some ways, but is superfluous or irrelevant in
event/history log. Use of the DCS in combination with others. How can Management decide what is important?
electronic batch records (EBR) would provide more
granularity to exception reporting through viewing Getting the control strategy right
exceptions within context of the batch. Use of EBR in an The first step is to make sure that the process control
integrated MES would provide the ability to reconcile strategy for the facility is appropriate, and that process,
and/or integrate data with separate deviation manage- reporting, and quality parameters are identified and
ment systems. And finally, Process Analytical Technol- assigned. These include input parameters such as equip-
ogy (PAT) in-line instrumentation and modeling can be ment settings (e.g. agitation rate) and output parameters
 P r o c e s s Va l i d at i o n, A u t o m at i o n a n d C o n t r o l

(e.g. pH achieved based on assignment and use of the number)? If so, are those fields appropriate for receiving data
input parameter agitation rate). How critical are each of from an MES to have successful reconciliation?
these? Not all parameters are critical. The assignment of
Critical Quality Attributes should be based on a careful Program definition and procedures
risk assessment with risk to patient as the focus. Many companies already have strong process monitoring
With the key parameters identified and the critical ones initiatives, but not all of them tie back to the validation
called out, it’s time to define the statistical process control program and have the appropriate quality oversight; the
ranges based on live data (batches performed to provide goal is to combine process monitoring data with the valida-
enough sample size to provide assurance of control). tion program to comply with CPV requirements and derive
Control ranges should be tighter than acceptance criteria. business and process benefits. The CPV program should be
In those key areas where analytical methodologies maintained in alignment with the master validation plans
already exist, the next step is to implement PAT for each site and/or process. Note again, this is recom-
methods, incrementally. Note that much more data will mended for products after the initial validation has been
be generated from the PAT methods than from manual done, so it can be implemented for a company’s current
laboratory methods. Where there may have been only commercial products once the appropriate reporting strat-
one data point in a laboratory for a particular attribute, egy is developed. In addition, if the output of the monitor-
there might be hundreds of data points with a PAT ing program is currently provided only to the technical
method. What is the appropriate response to that? staff, the audience list should be updated and a formal
What is the appropriate sample frequency? review program instituted to include the Quality Unit and
The quality of the other validation
raw materials also has Combine process monitoring data roles. The CPV pro-
an effect on product with the validation program to gram then becomes
quality. What are the a robust mechanism
comply with CPV requirements and
key attributes, whether for identification
received from vendor derive business/process benefits. of improvements
testing on a certificate and obtaining
of analysis (COA) or tested on-site as part of release collective agreement on implementation of changes.
process? It’s important to evaluate current data collection The formal program should identify the frequency for
methods to ensure these key attributes are funneled to the reporting and identify the mechanisms for investigation
appropriate place for analysis. and follow-up. This should include the definition of the
audience for reporting, as well as oversight by the Quality
Deviation monitoring Unit. Typically, the process, roles and responsibilities would
Earlier we mentioned deviation monitoring systems. It’s be defined in a Standard Operating Procedure under GMP
important to make sure that deviation monitoring is kept requirements to ensure consistent process and training.
in a single system, rather than maintaining bits of data Where real-time data from DCS, MES and/or PAT
scattered across systems. Keeping all the data together cannot be employed or a company is not ready to make
can help to ensure that all excursions are identified and that leap, it is still possible to achieve CPV through
that the associated data can be trended, so that reports incorporating laboratory data into the program. The
can be generated from a single source. This requires a frequency of reporting may differ based on lag time
system with the flexibility to report issues and investigate. to receive results, but analysis and improvement
The decision to launch an investigation should be predi- opportunities may be derived on a real-time basis rather
cated on potential impact. than traditional batch review during release processes;
The deviation monitoring system should include the with additional benefit derived when data is evaluated
appropriate data fields and reporting mechanisms to batch to batch instead of evaluating only a single batch.
reduce the effort expended on data mining. The system In short, the drivers behind FDA’s most recent process
should be made available to the plant’s MES to allow for validation guidance represent good science and facilitate
real-time documentation and reconciliation of issues as continuous improvement on the part of suppliers;
manufacturing occurs. To make this possible, it’s important both are critically important factors to the needs of the
to evaluate the deviation process workflow. Does the system patients and the reputation of the industry. The guidance
require a minimum number of fields to be populated before challenges pharmaceutical manufacturers to achieve
a record can be saved (and therefore receive an identification Continued Process Verification.

Reprinted with permission from Pharmaceutical Manufacturing, September 2012. On the Web at www.pharmamanufacturing.com.
© PUTMAN. All Rights Reserved. Foster Printing Service: 866-879-9144, www.marketingreprints.com.